All work
PB-01 · Lead Engine
Halden Freight
Every quote request answered in two minutes
Logistics · 140 people
Halden’s sales desk was losing freight quotes overnight. An agent now reads each request, prices the lane from their rate sheet and sends a draft quote before the customer calls anyone else.
2 min
first reply · was 9 h
+31%
more quotes won
14 h
saved every week

Spec sheet
Client
Halden Freight
Industry
Logistics · 140 people
Unit
PB-01 · Lead Engine
Wired into
Web forms, Gmail, HubSpot, Google Sheets, Slack
Live in
3 weeks
Year
2026
The problem
Halden quotes road freight across Europe, and most requests arrive after 6 pm, when a planner in another time zone finally gets to their inbox. The sales desk worked through the backlog every morning, but by then a third of those shippers had already booked with whoever replied first.
The rates were never the issue. Every price lived in one shared rate sheet that the team had kept up to date for years. The slow part was reading each request, finding the lane and writing the same email forty times a day.
What we wired
An intake agent that reads web forms and email, and pulls out origin, destination, weight, dates and special handling.
A pricing step that looks up the lane in the existing rate sheet and adds the surcharges the desk already applies.
A draft quote in Halden’s own template, sent within two minutes, with the deal logged in HubSpot.
Anything unusual, such as hazardous goods or a new country, goes to the desk in Slack instead of to the customer.
What changed
First replies dropped from nine hours to under two minutes, and the share of quotes Halden wins went up by almost a third in the first quarter. The desk now spends its mornings on the large contracts that need a person, and approves the routine drafts from their phones.
Transcript · client
2026
We used to start every morning with forty unanswered quote requests. Now we start with forty drafts to approve.
Marit Halden
Commercial Director, Halden Freight
Next build
Juniper Dental
62% of patient questions solved before the front desk sees them
